Python is known as a slow programming language. It is nonetheless very popular in the scientific community, and is used to perform massive numerical computations. How can that be? In a word: NumPy.
NumPy, or Numerical Python, offers a very efficient way fo storing numerical arrays of arbitrary dimensions. This talk will cover some tips and tricks for an optimal use of NumPy array, including vectorization, broadcasting, ufuncs and strides.
(Video is in French.)